Veteran Status and Substance Use Among Homeless Men in Los Angeles' Skid Row Neighborhood

Purpose: Homeless individuals have higher morbidity and mortality rates than the general population with veterans being overrepresented. Research is contradictory regarding the impact veteran status has on substance abuse of homeless men. Current drug use research appears to include all drugs in one category instead of analyzing specific substances. Based on results from a previous study conducted in Los Angeles, we hypothesized that homeless male veterans have higher rates of alcohol, marijuana and cocaine use than non-veterans. Due to minimal research available regarding usage of crack, heroin, prescription drugs, and methamphetamines, we hypothesized homeless veterans are different from non-veterans regarding substance use. An interaction effect of veteran status and PTSD diagnosis was performed to determine a possible moderation effect of PTSD on substance use.  

Methods: The University of Southern California/RAND Men’s Homeless Survey was the dataset used. A stratified random sampling method was used to recruit 305 homeless men from 13 meal lines and shelters of the Skid Row area of Los Angeles. Data was collected through in-person interviews assessing substance use within the past six months for the substances listed above. Post Traumatic Stress Disorder (PTSD) was determined by administering the PC-PTSD; scores of 3 or higher indicated PTSD. Using SAS 9.3, logistic regressions were performed to analyze the impact of veteran status on substance use within the past 6 months. Substance use was dichotomized.

Results: The average age in this all-male sample was 45 with the majority being African American. Nearly 20% of the sample endorsed veteran status; 28% of veterans were veterans of Iraq and Afghanistan Wars. Data were weighted to ensure the men sampled from shelters and meal lines represented the population of homeless men in Skid Row. Controlling for demographic variables (age, marital status, race/ethnicity), trends at p< .10 were evident regarding veteran status and usage of marijuana (OR=.48, 95% CI= .21, 1.10), cocaine (OR=2.99, 95% CI=.87, 10.3), heroin (OR=.06, 95% CI= .00, 2.71), prescription drugs (OR=2.91, 95% CI=1.46, 6.82), and methamphetamines (OR=3.25, 95% CI= .86, 12.34). Correlations at p<.05 were found between PTSD and usage of cocaine (OR=3.48, 95% CI= 1.45, 8.33), crack (OR=2.36, 95% CI= 1.36, 4.09), and prescription drugs (OR=3.16, 95% CI= 1.46, 6.82). There were no interaction effects for veteran status and PTSD diagnosis on substance use.

Conclusions and Implications: Although veteran status and substance use were not significantly associated, several trends emerged. Our non-significant findings of substance use and the interaction effect of veteran status and PTSD may be attributed to the heterogeneity of the veteran subset as military experiences differ among various cohorts. Additional studies with larger subsets of veterans from distinct cohorts are needed to determine specific relationships. The significant associations between PTSD and use of cocaine, crack, and prescription drugs highlights the importance of addressing comorbidities in the homeless population. As veterans are not eligible for mental health care at the VA without a service-connected disability, better knowledge of substance use trends and comorbidities would assist community agencies with appropriate service delivery to meet their needs.